HP 49g Graphing Calculator - Manual (Page 173)

Fitting a model to a set of data
You can use the statistics application to calculate Pearson’s correlation
coefficient for bivariate data. The statistics application quantifies the
coiTelation between data in any two columns in a matrix. You can choose
a regression model to apply to the data to find the relationship, or you can
select the Best Fit option to allow the calculator to find the best
correlation from its library of fit types.
The following four regression models are available for selection:
• Linear fit
y = b + nix
® Logarithmic fit
y = b + mlnr
® Exponential fit
y = or \ny = \nb + mx
® Power fit
y = b.x™ or ln?y = ln6 + minx
To determine details of tire regression model that applies to your data, use
the following procedure:
1. Use the method described in “Starting an application and specifying
the data” on page 9-2 to open the Fit Data input form and to load the
data to analyze.
2. In the X-Col field, enter the column munber that holds the independent
variable values.
3. In the Y-Col field, enter the column number that holds the dependent
variable values.
4. Place the cursor in the Model field and press
CHOOS.
A choose list
containing the regression model options is displayed.
5. Select the regression model to apply to the data, or select Best fit to
apply the model that best fits the data.
6.
Press
OK
to calculate the regression details. The following items
appear in history.
- Item 1: the covariance value.
- Item 2: the correlation coefficient.
- Item 3: the regression fonnula.
Using statistics Page 9-5